RNA interference is the process of sequence-specific, post-transcriptional gene silencing in animals and plants initiated by double stranded (ds) RNA that is homologous to the silenced gene (Hammond, S. M. et al., 2000; Fire, A., 1999; Sharp, P. A., 2001). In particular, synthetic and endogenous siRNAs are known to direct targeted mRNA degradation (Hammond, S. M. et al., 2000; Elbashir, S. M. et al., 2001; Caplen, N. J. et al., 2001; Clemens, J. C. et al., 2000; Lipardi, C. et al., 2001; Elbashir, S. M. et al., 2001; Ui-Tei, K. et al., 2000).
This powerful genetic technology has usually involved injection or transfection of ds RNA in model organisms. RNA interference also is a potent inhibitor of targeted gene expression in a variety of organisms (Wianny, F. et al., 2000; Kennerdell, J. R. et al., 1998; Fire, A. et al., 1998; Oelgeschlager, M. et al., 2000; Svoboda, P. et al., 2000). Recent studies by several groups (Lipardi, C. et al., 2001; Sijen, T. et al., 2001) suggest that ds small interfering RNAs (siRNAs) are part of a riboprotein complex that includes an RNAse III-related nuclease (Dicer) (Bernstein, E. et al., 2001), a helicase family (Dalmay, T. et al., 2001; Cogoni, C. et al., 1999), and possibly a kinase (Nykanen, A. et al., 2001) and an RdRP (Lipardi, C. et al., 2001; Smardon, A. et al., 2000). The mechanism proposed by Lipardi et al. (Lipardi, C. et al., 2001) is that one of the siRNA oligomers (antisense to the target RNA) primes an RdRP, generating longer dsRNAs, which are then cleaved by the RNAse III activity into additional siRNA duplexes, thereby amplifying the siRNAs from the target template.
dsRNA ≧30 bp can trigger in mammalian cells interferon responses that are intrinsically sequence-nonspecific (Elbashir, S. M. et al., 2001). However, duplexes of 21-nucleotide (nt) siRNAs with short 3′ overhangs can mediate RNA interference in a sequence-specific manner in cultured mammalian cells (Elbashir, S. M. et al., 2001). Two groups have demonstrated that 19 to 21 base duplexes with 3′UU or TT overhangs can effectively elicit an siRNA response in mammalian cells (Elbashir, S. M. et al., 2001; Caplen, N. J. et al., 2001). However, one limitation to the use of siRNA as a therapeutic reagent in vertebrate cells is that short, highly defined RNAs need to be delivered to target cells, which thus far has been accomplished only by using synthetic, duplexed RNAs delivered exogenously to cells (Elbashir, S. M. et al., 2001; Caplen, N. J. et al., 2001).
The present invention overcomes at least the above limitation.